We are performing a systematic screening of the human genome in order to identify a gene responsible for manic depressive disorder in the Old Order Amish. We are utilizing a large, multigenerational Old Order Amish pedigree with many affected individuals. By using this multigenerational pedigree with a high incidence of a disease manifested by a consistent and definable phenotype, we reduce uncertainties in analysis that could otherwise be introduced as a consequence of genetic heterogeneity, variable mode of inheritance, phenocopies, and/or penetrance. Highly informative markers (RFLP, tandem repeats, dinucleotide repeats, etc.) scattered across the genome are used to obtain genotypings. Due to the uncertainties in ascertaining phenotype and because of the lack of any definitive biological marker for bipolar illness (manic-depressive disorder), the linkage analyses are performed using several clinical hierarchies. Associative analyses are also carried out. Computer simulations using specific models are used to demonstrate the power of analyses. Simulation and data analysis includes several models of inheritance and diagnostic hierarchies, as well as polygenic etiologies. As marker genotypings are accumulated, the diagnostic data is updated, and our collection of cell lines from both normal and affected family member is being expanded. The availability of longitudinal follow-up evaluation for family members and the high frequency of manic depressive illness in the pedigree make the old Order Amish pedigree a valuable resource in the search for the genetic loci involved in bipolar affective disorder.